This application builds on a current gene transfer protocol in which the neomycin resistance gene (NEO-R) is used as a marker to determine the source of relapse in neuroblastoma patients who undergo autologous bone marrow transplantation. The proposed research assesses the ability of autologous neuroblastoma cells expressing IL2 after gene transduction to generate an antitumor response in patients with relapsed/resistant disease. Cell lines are being prepared from patients presenting with stage C or D neuroblastoma before they receive "front-line" therapy. These cell lines are transduced with a vector containing both the NEO-R marker and the IL2 gene, after which G418 resistant cytokine-secreting sublines are selected for freezing. Approximately 80% of the patients from whom the lines are obtained will relapse within three years. At the time of failure, patients will receive two s.c. injections of thawed tumor cells one week apart, in a phase I a/b dose escalation study. Toxicity will be evaluated clinically and by biopsy of injection sites. Local immune responses will be investigated by immunofluorescence studies and PCR analysis of T cell receptor usage. Systemic immunity will be analyzed by measuring proliferative and cytotoxic effects against autologous and allogeneic neuroblastoma cell lines, and by detecting changes in T cell receptor usage. Clinical efficacy will be determined from evaluations of disease progression by bone marrow biopsy, and by imaging techniques including CT, MRI or isotope bone scans, as indicated. If this approach to immunotherapy is safe and immunologically efficacious, there would be justification for phase II testing of the same strategy in patients with advanced disease and with minimal residual disease. The approach could also be extended to other cytokines and cytokine combinations.